DESCRIPTION: Suicide and lesser forms of intentional self-harm behaviors produce devastating medical, social and economic costs. Self-harm is integrally related to depressive disorders and Borderline Personality Disorder. Selective Serotonin Reuptake Inhibitors (SSRIs), like escitalopram, are front-line pharmacological treatments for these disorders, putatively regulating depressed mood and reducing suicidality. However, data from case studies and retrospective meta-analyses of depression clinical trials is mixed, with some (but not all ) studies suggesting that during the first months of treatment, SSRIs may increase the risk of suicidal ideation in select individuals, particularly younger individuals. These post-hoc analyses, though informative, are based on studies that provide limited sampling of the self-harm domain. No study, to date, has implemented a direct prospective examination of the effects of early SSRI use on self-harm thoughts and behaviors using a multi-method measurement involving both the laboratory (standard self-aggression paradigm: SAP) and home environments (ecological momentary assessment: EMA). Also, no study has examined the influence of impaired 5-HT function and emotion dysregulation as moderators of outcome with escitalopram. The proposed randomized clinical trial will prospectively assess the impact of eight weeks exposure to SSRI treatment on self-harm ideation and behavior among a sample of 200 subjects with Borderline Personality Disorder and current major depression. After a one week single-blind placebo lead-in, participants will be randomly assigned double blind to either placebo or escitalopram for eight (8) weeks. The primary dependent variable will be EMA of self-harm ideation and behavior obtained several times each day. Self-harm will also be assessed using a laboratory analogue task (SAP) at baseline and again after the eight week trial. Age will be evaluated as a moderator of SSRI response. 5-HT dysfunction and emotion dysregulation will be evaluated as candidate moderators of SSRI response. 5-HT functioning will be assessed using psychophysiological (loudness dependence of the auditory evoked potential: LDAEP) and genetic (5-HT transporter promoter polymorphism: 5-HTTLPR) markers. Measures of emotion dysregulation will include trait aggression, impulsivity and socioemotional information processing. At the conclusion of the eight-week randomized trial, all participants will receive eight weeks of escitalopram administered single-blind, with continued EMA and other assessment.